The AMD Radeon RX 590 Review, feat. XFX & PowerColor: Polaris Returns (Again)
by Nate Oh on November 15, 2018 9:00 AM ESTCompute & Synthetics
Shifting gears, we'll look at the compute and synthetic aspects of the RX 580. As this is the 3rd time we're seeing 'big' Polaris, there's nothing new to discuss, and with no other design changes we only expect to see the impact of increased clockspeeds.
Beginning with CompuBench 2.0, the latest iteration of Kishonti's GPU compute benchmark suite offers a wide array of different practical compute workloads, and we’ve decided to focus on level set segmentation, optical flow modeling, and N-Body physics simulations.
Moving on, we'll also look at single precision floating point performance with FAHBench, the official Folding @ Home benchmark. Folding @ Home is the popular Stanford-backed research and distributed computing initiative that has work distributed to millions of volunteer computers over the internet, each of which is responsible for a tiny slice of a protein folding simulation. FAHBench can test both single precision and double precision floating point performance, with single precision being the most useful metric for most consumer cards due to their low double precision performance.
Next is Geekbench 4's GPU compute suite. A multi-faceted test suite, Geekbench 4 runs seven different GPU sub-tests, ranging from face detection to FFTs, and then averages out their scores via their geometric mean. As a result Geekbench 4 isn't testing any one workload, but rather is an average of many different basic workloads.
We'll also take a quick look at tessellation performance.
Finally, for looking at texel and pixel fillrate, we have the Beyond3D Test Suite. This test offers a slew of additional tests – many of which use behind the scenes or in our earlier architectural analysis – but for now we’ll stick to simple pixel and texel fillrates.
Facebook
Twitter
RSS
136 Comments
View All Comments
neblogai - Thursday, November 15, 2018 - link
For me, and many other people 2080 or 2080Ti might as well not exist. I'd never buy any GPU priced like that. In fact, I'd never even buy a €300+ card. Also- power use is not an issue- at 200W+, it is easy to cool, not noisy, and savings in electricity cost from having a GTX1060 would be only €5-10 per year (at 20 hour gaming per week, which I do not achieve). And in a sub-€300 market- nVidia has not offered anything for 2.5 years. So it is certainly better deal to buy a faster RX590 and get a €180 AAA game bundle for free, compared to similarly priced, slower GTX1060. There are other reasons as well, like Freesync, and futureproofing like 8GB of VRAM and better driver support (because nVidia will be moving toward doing optimizations for 20xx series).eva02langley - Thursday, November 15, 2018 - link
It is a polaris card. Everybody knew what to expect. I am not sure what you are talking about.Navi is due next year and the contender will be the 2070 RTX. I am going to speculate a tag price of 300$ which is a whole 200-300$ less than a RTX.
mapesdhs - Thursday, November 15, 2018 - link
Xex360, AMD is not trying to compete because just don't buy them, so what's the point? There market and brand awareness simply isn't there to support a high end product stack atm, not until gamers stop being so irrational and actually buy AMD when they are objectively the more sensible option, whether based on price, performance or some combination of metrics.Cooe - Thursday, November 15, 2018 - link
"but now built on GlobalFoundries' 12nm process"Sorry, Nate, but ya got that one wrong. Polaris 30 is being fabbed at TSMC, just like the rest of AMD's GPUs.
eva02langley - Thursday, November 15, 2018 - link
... sure....porcupineLTD - Thursday, November 15, 2018 - link
Where are people getting this from?Ryan Smith - Thursday, November 15, 2018 - link
Note that we currently have no information cooroborating the use of TSMC. And indeed it seems incredibly unlikely given AMD's WSA, and the fact that Polaris 30 has the exact same die size as Polaris 10. TSMC and GF's 12nm processes are not identical, porting a chip to TSMC would given you different geometry dimensions.lmcd - Friday, November 16, 2018 - link
Just like the rest of AMD's...https://www.anandtech.com/show/10446/the-amd-radeo...
GPUs. Right.
maroon1 - Thursday, November 15, 2018 - link
What a jokeOnly 12% faster but also cost and consume more power than RX580 (and 108watt more than GTX 1060 at load)
It adds noting new. It just fills the big gap between RX580 and vega56
dr.denton - Saturday, November 17, 2018 - link
"It adds noting new. It just fills the big gap between RX580 and vega56"Literally all it was ever intended to do. Polaris is a 2 year old mid range GPU design, how do you expect anything revolutionary from that?